| Summary: | sci-mathematics/octave - gnuplot communication results in wrong data displayed | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Arrigo Marchiori <ardovm> |
| Component: | Current packages | Assignee: | Gentoo Science Mathematics related packages <sci-mathematics> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | major | CC: | fauli, sci, ulm |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
|
Description
Arrigo Marchiori
2008-10-17 09:53:57 UTC
Can you somewho extract the generated Gnuplot code from Octave? (In reply to comment #1) > Can you somewho extract the generated Gnuplot code from Octave? Ok, that's what I've done (transcripts follow): 1. run octave 2. enter the commands: gnuplot_binary = "cat" x = [0:100]; y = sin(x); plot(x, y) 3. octave shows: set data style lines set nologscale set nopolar pl '/tmp/oct-wvb00G' u 1:2 t "line 1" 4. enter the command: plot(x, t) 5. octave shows: set nologscale set nopolar pl '/tmp/oct-0LGsRp' u 1:2 t "line 1" 6. open gnuplot and enter the first commands. Each command issues a warning but, in the end, a window is displayed, showing a nice sine wave. 7. in the same gnuplot session, enter the second set of commands. Gnuplot complains with warning at the "set" commands, and then displays the wrong data: a flat line, together with the warning message about the y range. The two data files that Octave wants to feed into gnuplot contain the same data except the very first line, that is a comment. $ diff /tmp/oct-wvb00G /tmp/oct-0LGsRp 1c1 < # name: /tmp/oct-wvb00G --- > # name: /tmp/oct-0LGsRp $ ---- Transcript of octave session ---- octave:1> gnuplot_binary = "cat" gnuplot_binary = cat octave:2> x = [0:100]; octave:3> y = sin(x); octave:4> plot(x, y) set data style lines set nologscale set nopolar pl '/tmp/oct-wvb00G' u 1:2 t "line 1" octave:5> plot(x, y) set nologscale set nopolar pl '/tmp/oct-0LGsRp' u 1:2 t "line 1" octave:6> ---- Transcript of gnuplot session ---- gnuplot> set data style lines ^ warning: deprecated syntax, use "set style data" gnuplot> set nologscale ^ warning: deprecated syntax, use "unset" gnuplot> set nopolar ^ warning: deprecated syntax, use "unset" gnuplot> pl '/tmp/oct-wvb00G' u 1:2 t "line 1" gnuplot> set nologscale ^ warning: deprecated syntax, use "unset" gnuplot> set nopolar ^ warning: deprecated syntax, use "unset" gnuplot> pl '/tmp/oct-0LGsRp' u 1:2 t "line 1" Warning: empty y range [0:0], adjusting to [-1:1] gnuplot> Thank you for your help! Have you tried if octave-3* works for you? It has a much improved interface to gnuplot and I recommend that you give it a try. I will probably try to move octave-3* into stable sometime in the not too far future, but there are a lot of dependencies that need to go stable first hence this will take some time. Best, Markus (In reply to comment #3) > Have you tried if octave-3* works for you? I just tested it. It seems to be working fine! In order to compile it, though, I had to unmask sci-mathematics/glpk and disable the "mysql" USE flag for it, otherwise it didn't compile. I can post another bug report for that, if you wish. Thank you for your help! (In reply to comment #4) > (In reply to comment #3) > > Have you tried if octave-3* works for you? > > I just tested it. It seems to be working fine! Good! > > In order to compile it, though, I had to unmask sci-mathematics/glpk and > disable the "mysql" USE flag for it, otherwise it didn't compile. I can post > another bug report for that, if you wish. > Please do (if there isn't one already) so we can keep track of it. Thanks in advance. Best, Markus (In reply to comment #5) > (In reply to comment #4) > > (In reply to comment #3) > > > Have you tried if octave-3* works for you? > > > > I just tested it. It seems to be working fine! > > Good! Yes: now I can work again with Octave! The problem is not fixed, technically speaking... but I'm happy with this solution :-) > > > > In order to compile it, though, I had to unmask sci-mathematics/glpk and > > disable the "mysql" USE flag for it, otherwise it didn't compile. I can post > > another bug report for that, if you wish. > > > > Please do (if there isn't one already) so we can keep track of it. Well... it didn't work yesterday but it worked today. :-$ So, nothing to report about that. Anyway, it was /usr/include/mysql/my_global.h pointing to a non-existant include file <asm/atomic.h>. But, I repeat, it worked for me just now. Thank you again! (In reply to comment #6) > (In reply to comment #5) > > (In reply to comment #4) > > > (In reply to comment #3) > > > > Have you tried if octave-3* works for you? > > > > > > I just tested it. It seems to be working fine! > > > > Good! > > Yes: now I can work again with Octave! > > The problem is not fixed, technically speaking... but I'm happy with this > solution :-) > > > > > > > In order to compile it, though, I had to unmask sci-mathematics/glpk and > > > disable the "mysql" USE flag for it, otherwise it didn't compile. I can post > > > another bug report for that, if you wish. > > > > > > > Please do (if there isn't one already) so we can keep track of it. > > Well... it didn't work yesterday but it worked today. :-$ So, nothing to report > about that. Anyway, it was /usr/include/mysql/my_global.h pointing to a > non-existant include file <asm/atomic.h>. But, I repeat, it worked for me just > now. > The mysql clash reported under bug #254136. octave-2 bug which we do not care anymore. |
